Acoustic tomography involves sending a sonic signal across a measurement area and timing how long it takes to travel the distance. A number of transducers are situated around the measurement area to achieve this. Since the absolute sonic speed is affected by temperature and wind speed, so is the time of flight. It is possible to reconstruct the temperature and wind velocity from the collection of time-of-flight data.
This package gives an example of using radial basis function networks to allow for a linear solution to the inverse problem. The example data is from a Karman vortex street behind a cylinder.
See http://blog.nutaksas.com for more details.
Removed GPL per Mathworks' requirements.